Attractive people Assembling loose-limbed models using non-parametric belief propagation.pdf

Attractive People: Assembling Loose-Limbed Models using Non-parametric Belief Propagation Leonid Sigal Department of Computer Science Brown University Providence, RI 02912 ls@ Michael Isard Microsoft Research Silicon Valley Mountain View, CA 94043 misard@ Benjamin H. Sigelman Department of Computer Science Brown University Providence, RI 02912 bhsigelm@ Michael J. Black Department of Computer Science Brown University Providence, RI 02912 black@ Abstract The detection and pose estimation of people in images and video is made challenging by the variability of human appearance, the complexity of natural scenes, and the high dimensionality of articulated body mod- els. To cope with these problems we represent the 3D human body as a graphical model in which the relationships between the body parts are represented by conditional probability distributions. We formulate the pose estimation problem as one of probabilistic inference over a graphi- cal model where the random variables correspond to the individual limb parameters (position and orientation). Because the limbs are described by 6-dimensional vectors encoding pose in 3-space, discretization is im- practical and the random variables in our model must be continuous- valued. To approximate belief propagation in such a graph we exploit a recently introduced generalization of the particle filter. This framework facilitates the automatic initialization of the body-model from low level cues and is robust to occlusion of body parts and scene clutter. 1 Introduction Recent approaches to person detection and tracking exploit articulated body models in which the body is viewed as a kinematic tree in 2D [14], 2.5D [16, 23], or 3D [2, 5, 6, 19, 21] leading to a parametric state-space representation of roughly 25–35 di- mensions. The high dimensionality of the resulting state-space has motivated the devel- opment of specialized stochastic search algorithms that either exploit the highly redundant dynamics of typical human motion